Best OCR in uipath

Which OCR is best suited for text from image extraction?
and Tell which activity is to use for extraction from image
1)Get OCR text
2)Load image
3)Text detection

Hi @AbarnaKalaiselvam

The answer will be Get OCR Text.
Explanation:

  1. Get OCR Text: This is the activity you should use to extract text from an image. It supports various OCR engines and allows you to specify the OCR engine and other relevant parameters.
  2. Load Image: This activity is used to load an image from a file or a specified path. It’s not directly related to OCR, but you would typically use it to provide the image as input to OCR activities.
  3. Text Detection: This activity might be used for identifying regions of text within an image, which could be useful if you’re working with images that contain multiple blocks of text. However, the specific activity names and functionalities might vary based on updates in UiPath’s libraries after my last knowledge update.

Hope it helps!!

Hi @AbarnaKalaiselvam,

1)Get OCR Text : This is the activity you should use to extract text from an image. It supports various OCR engines and allows you to specify the OCR engine and other relevant parameters.

Cheers

Which ocr is best for image extraction?

1)Omnipage
2)Tesseract OCR
3)Google cloud vison or
any other

Hi @AbarnaKalaiselvam

Tesseract OCR and Omni Page OCR works fine with extraction. Try with Tesseract OCR for extraction first and if your result is not accurate you can go with OmniPage OCR by downloading the package UiPath.OmniPage.Activities.
Read this explanation for better understanding

  1. OmniPage:OmniPage is a commercial OCR software that’s not natively integrated into UiPath. However, you can use UiPath’s activities to interact with external applications, including OmniPage. You could design a workflow that involves the following steps:
    a. Start OmniPage: Use UiPath’s “Start Process” activity to launch the OmniPage application.
    b. Load Image and Extract Text: Within OmniPage, you would load the image using its user interface and then perform OCR to extract text. Depending on OmniPage’s features, you might be able to automate these actions using UiPath’s “Click” and “Type Into” activities to simulate mouse clicks and keyboard inputs.
    c. Retrieve Text: After text extraction in OmniPage, you can use UiPath to retrieve the extracted text from OmniPage’s user interface or a specific output location.

  2. Tesseract OCR:Tesseract OCR can be integrated into UiPath using custom activities or scripts. UiPath does not natively include Tesseract OCR activities, but you can create a custom workflow like this:
    a. Install Tesseract: Set up Tesseract OCR on your machine or a server that UiPath can access.
    b. Invoke Code: Use the “Invoke Code” activity in UiPath to execute a custom script that uses Tesseract to perform OCR on the provided image. You would need to write code in a language like Python to interact with the Tesseract API.
    c. Retrieve Text: The custom script should output the extracted text, which you can then retrieve using UiPath’s standard variables.

  3. Google Cloud Vision OCR:Google Cloud Vision OCR is a cloud-based service provided by Google. UiPath offers a dedicated activity to interact with Google Cloud services, making integration relatively straightforward:
    a. Set Up Google Cloud Account: Sign up for a Google Cloud account and set up the Cloud Vision API.
    b. Use “Google Cloud Vision OCR” Activity: In UiPath, use the “Google Cloud Vision OCR” activity to send the image to the Google Cloud service for text extraction. You’ll need to provide the necessary credentials and specify the features you want to use.
    c. Retrieve Text: The activity will return the extracted text, which you can then use in your UiPath workflow.

Hope it helps!!

Thanks for all of your replies.

If i’m using Get OCR text activity, How to select multiple set of images from folders.
Is there any way to dynamically for set of images from folders.

Hi @AbarnaKalaiselvam

If you find solution for your query please mark it as solution to close the loop.

Happy Automation

Regards

@AbarnaKalaiselvam

you need to place in the loop by making it dynamic

cheers

ok, but i have to indicate the screen for the particular image right?
then how to do in loop?

@AbarnaKalaiselvam

you can make the selector dynamic so that it will works for multiple set of images

cheers

can u explain how to make the selectors as dynamic?

Hi @AbarnaKalaiselvam

Can you tell me your usecase clearly? If you have different sets of Images and want to extract the text then you can do as below

  1. Use assign activity and declare a string variable, let’s say arrFiles now write as Directory.GetFiles(FolderPath)
  2. Then you can use ForEach loop and write arrFiles, Now you can place your Get OCR text activity inside the For Each loop and pass the fileName which is currentItem to the Get OCR activity

that will be very basic approach, if you are usecase is more complicated then this Get OCR text will not work and you need to work on Document Understanding

Hope this may help you

Thanks,
Srini

@AbarnaKalaiselvam

based on the images you need to make the selector dynamice

like title and aaname attributes making dynamic

if you want any more share any images

cheers

1 Like

Hi @AbarnaKalaiselvam ,
Get OCR

1 Like

@AbarnaKalaiselvam

can you send the selectors by indicating 2 images

so that i can make dynamic and send you the selector

Extract the text from an images dynamically.

Dynamically means from an input folder i need to extract all the image text from the input image folder.


@AbarnaKalaiselvam

in both the images use element level selector as

in the top level selector make title=‘*’

let me know is it working or not

23.8.0-beta.13817+Branch.release-v23.8.0.Sha.120cd3ec54f082511b48105e2cfb52f8897346a8

Source: Get OCR Text

Message: Could not find the UI element corresponding to this selector:
[1]
[2]

Search failed at selector tag:
[1]

The closest matches found are:
[33%]

Exception Type: UiPath.Core.SelectorNotFoundException

UiPath.Core.SelectorNotFoundException: Could not find the UI element corresponding to this selector:
[1]
[2]

Search failed at selector tag:
[1]

The closest matches found are:
[33%] at UiPath.Core.Activities.GetOCRText.EndExecute(NativeActivityContext context, IAsyncResult result)
at UiPath.Core.Activities.ObsoleteAsyncNativeActivity.BookmarkResumptionCallback(NativeActivityContext context, Bookmark bookmark, Object value)
at System.Activities.Runtime.BookmarkCallbackWrapper.Invoke(NativeActivityContext context, Bookmark bookmark, Object value)
at System.Activities.Runtime.BookmarkWorkItem.Execute(ActivityExecutor executor, BookmarkManager bookmarkManager)